Differentiation of Bifidobacterium longum subspecies longum and infantis by quantitative PCR using functional gene targets

Background Members of the genus Bifidobacterium are abundant in the feces of babies during the exclusively-milk-diet period of life. Bifidobacterium longum is reported to be a common member of the infant fecal microbiota. However, B. longum is composed of three subspecies, two of which are represented in the bowel microbiota (B. longum subsp. longum; B. longum subsp. infantis). B. longum subspecies are not differentiated in many studies, so that their prevalence and relative abundances are not accurately known. This may largely be due to difficulty in assigning subspecies identity using DNA sequences of 16S rRNA or tuf genes that are commonly used in bacterial taxonomy. Methods We developed a qPCR method targeting the sialidase gene (subsp. infantis) and sugar kinase gene (subsp. longum) to differentiate the subspecies using specific primers and probes. Specificity of the primers/probes was tested by in silico, pangenomic search, and using DNA from standard cultures of bifidobacterial species. The utility of the method was further examined using DNA from feces that had been collected from infants inhabiting various geographical regions. Results A pangenomic search of the NCBI genomic database showed that the PCR primers/probes targeted only the respective genes of the two subspecies. The primers/probes showed total specificity when tested against DNA extracted from the gold standard strains (type cultures) of bifidobacterial species detected in infant feces. Use of the qPCR method with DNA extracted from the feces of infants of different ages, delivery method and nutrition, showed that subsp. infantis was detectable (0–32.4% prevalence) in the feces of Australian (n = 90), South-East Asian (n = 24), and Chinese babies (n = 91), but in all cases at low abundance (